Next Article in Journal
Zingiber mioga Extract Improves Moisturization and Depigmentation of Skin and Reduces Wrinkle Formation in UVB-Irradiated HRM-2 Hairless Mice
Previous Article in Journal
The Influence of the Modernization of the City Sewage System on the External Load and Trophic State of the Kartuzy Lake Complex
Previous Article in Special Issue
Semantic Mapping with Low-Density Point-Clouds for Service Robots in Indoor Environments
Article

Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy

1
Département Mathématiques Informatique Statistique, Université Bretagne Sud, Lab-STICC, UBL, 56321 Lorient, France
2
IMT Atlantique, Lab-STICC, UBL, 29238 Brest, France
3
Flowers Team, U2IS, ENSTA Paris, Institut Polytechnique de Paris & Inria, 91120 Palaiseau, France
*
Author to whom correspondence should be addressed.
Academic Editor: Antonio Bandera
Appl. Sci. 2021, 11(3), 975; https://doi.org/10.3390/app11030975
Received: 16 December 2020 / Revised: 16 January 2021 / Accepted: 17 January 2021 / Published: 21 January 2021
(This article belongs to the Special Issue Cognitive Robotics)
In open-ended continuous environments, robots need to learn multiple parameterised control tasks in hierarchical reinforcement learning. We hypothesise that the most complex tasks can be learned more easily by transferring knowledge from simpler tasks, and faster by adapting the complexity of the actions to the task. We propose a task-oriented representation of complex actions, called procedures, to learn online task relationships and unbounded sequences of action primitives to control the different observables of the environment. Combining both goal-babbling with imitation learning, and active learning with transfer of knowledge based on intrinsic motivation, our algorithm self-organises its learning process. It chooses at any given time a task to focus on; and what, how, when and from whom to transfer knowledge. We show with a simulation and a real industrial robot arm, in cross-task and cross-learner transfer settings, that task composition is key to tackle highly complex tasks. Task decomposition is also efficiently transferred across different embodied learners and by active imitation, where the robot requests just a small amount of demonstrations and the adequate type of information. The robot learns and exploits task dependencies so as to learn tasks of every complexity. View Full-Text
Keywords: curriculum learning; continual learning; hierarchical reinforcement learning; interactive reinforcement learning; imitation learning; multi-task learning; active imitation learning; hierarchical learning; intrinsic motivation curriculum learning; continual learning; hierarchical reinforcement learning; interactive reinforcement learning; imitation learning; multi-task learning; active imitation learning; hierarchical learning; intrinsic motivation
Show Figures

Figure 1

MDPI and ACS Style

Duminy, N.; Nguyen, S.M.; Zhu, J.; Duhaut, D.; Kerdreux, J. Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy. Appl. Sci. 2021, 11, 975. https://doi.org/10.3390/app11030975

AMA Style

Duminy N, Nguyen SM, Zhu J, Duhaut D, Kerdreux J. Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy. Applied Sciences. 2021; 11(3):975. https://doi.org/10.3390/app11030975

Chicago/Turabian Style

Duminy, Nicolas, Sao M. Nguyen, Junshuai Zhu, Dominique Duhaut, and Jerome Kerdreux. 2021. "Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy" Applied Sciences 11, no. 3: 975. https://doi.org/10.3390/app11030975

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop